Acton¶

StratoWeave is written in the Acton Programming Language. The Acton guide provides comprehensive documentation on the language's syntax, features, and best practices. Try Ask Acton to get quick answers to specific questions.

As a Python developer, you will find a syntax that is instantly familiar, but there's also a few key differences to be aware of. The most obvious one is that Acton is a compiled and statically typed language. Yet, thanks to its powerful type inference, you can mostly write code without explicit type annotations and still get the benefits of static typing.

Optional types¶

In Acton's type system, a type can be either optional or non-optional. An optional type can be None or a value of the specified type, while a non-optional type must always have a value. In your StratoWeave transforms, the data nodes will be optional or non-optional based on how they are defined in the YANG model.

Consider this dummy YANG snippet.

list toaster {
  key name;
  sw:transform dummy.foo.Toaster;

  leaf name {
    type string;
  }
  leaf model {
    type string;
  }
  leaf capacity {
    type uint32;
    mandatory true;
  }
  leaf power {
    type uint32;
    default 11;
  }
}

• name because it's a key leaf
• capacity because it's a mandatory leaf
• power because it has a default value

The model property will be optional, however, which means that we must ensure it is not None before we can do string operations on it. Assign its value to a local variable and check for is not None before using it.

class Toaster(base.Toaster):
    def transform(self, i):
        print(i.name.upper())
        print(i.capacity + 1)
        print(i.power * 2)

        model = i.model # optional type, must check for None before using
        if model is not None:
            print(model.upper())

Failure to check for is not None on an optional type before using it will result in a compile-time error. While seemingly inconvenient at first, this strong typing helps catch potential bugs early and ensures that your transform logic is robust against missing or incomplete input data.

If you find that you have to write a lot of is not None checks that do not seem to add value, it may be a sign that the YANG model should be updated to make that node mandatory or give it a default value. Particularly for lower-layer transforms, it's often better to have a more strictly defined input schema to avoid having to write defensive code in the transform method.